It is relevant to note two conditions affecting the economic interpretation of the estimates. The first is that the values of traits are confined to those environments where disease and insect pressure is greatest. The second is that the trait values included in this study are incomplete. A new variety may have several traits and each has value. Since this study only covered two insect resistance and three disease resistance traits, it probably underestimates the full value of all traits.
Consider first the evidence regarding loss reduction. Specification (1) indicates that if all varieties had BPH resistance, losses from this pest would be reduced by 2%. Approximately the same can be said for GM resistance. In actuality only 60% of the varieties have BPH resistance and roughly 40% have GM resistance.
Thus by these estimates actual losses are only about 1% lower because of these two traits. But if we consider other insect pests and a further expansion of trait area, we could conclude that conventional plant breeding has reduced crop losses by 3-5% (considering these two insect pests to represent 25-33% of insect problems). There appears to be future potential for another 3-5% reduction if biotechnology methods enable a more complete incorporation of insect resistance traits.
For disease resistance traits, the evidence is less clear. Only rice Tungro resistance shows indication of loss reduction, and that is only by 0.3%. Even with some expansion to other diseases, it is difficult to say that disease resistance has contributed much more than 1% to crop-loss reduction to date.
The pesticide use estimates from specification 1 in Table 12.3 indicate that the total set of traits reduces pesticide use by 20% (the sum of coefficients is 4570, which is 80% of mean pesticide use). This amounts to roughly 1% of crop value.
Finally, the TFP equation can be utilized to calculate trait values. If we leave in the negative value for BPH, we obtain a coefficient of 0.46 for insect resistance and a combined coefficient of 0.53 for disease resistance. Multiplying these coefficients by adoption levels, these estimates imply that TFP indexes (yields) are higher (average costs are lower) by roughly 11% because of insect resistance and by 3% or so because of disease resistance traits.
The TFP-based estimates are higher than the combined crop-loss and pesticide estimates. With an expansion factor to cover other diseases and insects, the TFP evidence suggests that 15% of current TFP levels is due to these five traits. The generation 3 evidence (specification 3) indicates a 25% generation 3 gain. This is more than double the contributions suggested by the crop-loss and pesticide reduction estimates.
These estimates, however, can be reconsidered by noting that TFP (yields) may incorporate a synergistic effect, i.e. the sum is greater than the parts - in this case greater than the crop-loss parts.
It may thus be reasonable to conclude that, to date, rice yields in Indonesia are roughly 15% higher because of these traits, and that with synergism they may be 25% higher. It should be noted that this synergism is really due to quantitative trait improvement. Conventional plant breeding methods have allowed considerable gains to be realized in Indonesia and more are in the offing. This chapter has not directly considered ecological stress traits, although these may be captured in the TFP estimates, nor have grain quality traits been considered. Further research is required to more adequately address these issues.
Was this article helpful?